Proposta de um algoritmo eficiente baseado em busca tabu e representação nó-profundidade para a restauração de redes de distribuição de energia elétrica
Carregando...
Data
2016-08-05
Autores
Mathias-Neto, Waldemar Pereira [UNESP]
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
Os modernos sistemas aéreos de distribuição de energia elétrica são projetados para operar com altos índices de confiabilidade. Todavia, interrupções de fornecimento podem ocorrer a qualquer momento e causadas pelos mais diversos fatores, tais como, vendavais, árvores, ou ainda, colisão de veículos com a rede. Portanto, para garantir o mínimo impacto destas interrupções nos índices de confiabilidade da rede, áreas sem fornecimento devem ser reenergizadas tão rá- pido quanto possível. Este processo de reenergização do sistema é comumente denominado restauração. O principal objetivo do processo de restauração é restabelecer o maior número de cargas, no menor intervalo de tempo possível, por meio de uma sequência de aberturas e fechamentos de chaves de manobras. A existência de um conjunto de chaves seccionadoras estrategicamente posicionadas na rede permite que a estrutura malhada dos sistemas de distribuição opere com topologias radiais. Portanto, a execução de uma sequência de manobras é capaz de alterar a topologia radial da rede e restabelecer as cargas previamente sem fornecimento de energia. Este trabalho propõe uma nova metodologia para solução do problema de restauração de sistemas de distribuição de energia elétrica baseado no uso conjunto da meta-heurística de busca tabu, a representação nó-profundidade (RNP) e seus operadores PAO e CAO. Os operadores da RNP são empregados para originar a vizinhança da meta-heurística de busca tabu. Um novo operador foi introduzido e alterações nos demais operadores foram realizadas para possibilitar soluções factíveis em sistemas com alto carregamento. O problema é formulado através de um modelo não linear inteiro misto e considera a minimização dos custo da interrupção não programada como objetivo do problema. Os limites operacionais da rede são avaliados durante o intervalo previsto para a rede permanecer no estado restaurativo. Estes limites caracterizam as restrições do problema. Ao modelo matemático foi incluído a geração distribuída e cargas remotamente controladas como ferramentas de apoio à restauração. Os geradores distribuídos são modelados através de curvas horárias de geração, em função de sua tecnologia, e as cargas são modeladas através de curvas que consideram o perfil de consumo. Adicionalmente, cargas controladas termostaticamente também foram incluídas ao modelo. Por fim, a técnica de solução proposta é avaliada através de um sistema teste baseado no sistema de distribuição IEEE de 37 barras em oito diferentes cenários. Os resultados obtidos indicam a viabilidade da metodologia para a solução do problema de restauração de sistemas de distribui- ção.
The modern overhead power distribution systems are designed to operate with high reliability indices. However, power outages may occur at any time and caused by several factors, such aswindstorms, trees or vehicle collision with the network. Therefore, to ensure minimal impact of these interruptions in network reliability indices, out-of-service areas, which are not affected by the permanent fault, should be re-energized as fast as possible. The process of system reenergizing is called restoration. The main purpose of restoration process is to restore the largest amount of load, as fast as possible, through a sequence of switching actions. The existence of several switchgears, strategically allocated in the network, allows the meshed distribution systems work with radial topologies. Therefore, performing a sequence of maneuvers, automatically or manually, may change the radial grid topology and restore the loads previously without a power supply. This work proposes a new methodology to solve the distribution system restoration problem based on the joint use of tabu search meta-heuristic, the node-depth encoding (NDE) and its operators (PAO and CAO). The NDE operators are used to buid the neighborhood of tabu search meta-heuristic. A new operator was introduced and changes in other operators were carried out to allow feasible solutions in systems with heavy loading. The proposed mathematical model minimize the cost of unscheduled outage aim of the problem. The problem of restrictions are characterized by the operating limits of the network and evaluated periodically while the network remains in the restorative state. Additionally, the use of distributed generation and controlled remotely loads are considered to support the restoration tools. The problem is formulated as a nonlinear mixed-integer model. The main goal is minimize the cost of unscheduled interruption. The operating limits of the network are evaluated while it remains in the restorative state.. These limits characterize the constraints of the problem. At the mathematical model was included distributed generation and loads remotely controlled as support tools. The distributed generators are modeled by its hourly curves and the loads are modeled through curves with their consumption profile. Additionally, controlled thermostatically loads were also included into the mathematical model. Finally, the proposed solution is evaluated through a modified test system based on the IEEE 37-bus feeder test system. Eight different scenarios was evaluated. The results indicate the feasibility of the methodology to solve the distribution systems restoration problem.
The modern overhead power distribution systems are designed to operate with high reliability indices. However, power outages may occur at any time and caused by several factors, such aswindstorms, trees or vehicle collision with the network. Therefore, to ensure minimal impact of these interruptions in network reliability indices, out-of-service areas, which are not affected by the permanent fault, should be re-energized as fast as possible. The process of system reenergizing is called restoration. The main purpose of restoration process is to restore the largest amount of load, as fast as possible, through a sequence of switching actions. The existence of several switchgears, strategically allocated in the network, allows the meshed distribution systems work with radial topologies. Therefore, performing a sequence of maneuvers, automatically or manually, may change the radial grid topology and restore the loads previously without a power supply. This work proposes a new methodology to solve the distribution system restoration problem based on the joint use of tabu search meta-heuristic, the node-depth encoding (NDE) and its operators (PAO and CAO). The NDE operators are used to buid the neighborhood of tabu search meta-heuristic. A new operator was introduced and changes in other operators were carried out to allow feasible solutions in systems with heavy loading. The proposed mathematical model minimize the cost of unscheduled outage aim of the problem. The problem of restrictions are characterized by the operating limits of the network and evaluated periodically while the network remains in the restorative state. Additionally, the use of distributed generation and controlled remotely loads are considered to support the restoration tools. The problem is formulated as a nonlinear mixed-integer model. The main goal is minimize the cost of unscheduled interruption. The operating limits of the network are evaluated while it remains in the restorative state.. These limits characterize the constraints of the problem. At the mathematical model was included distributed generation and loads remotely controlled as support tools. The distributed generators are modeled by its hourly curves and the loads are modeled through curves with their consumption profile. Additionally, controlled thermostatically loads were also included into the mathematical model. Finally, the proposed solution is evaluated through a modified test system based on the IEEE 37-bus feeder test system. Eight different scenarios was evaluated. The results indicate the feasibility of the methodology to solve the distribution systems restoration problem.
Descrição
Palavras-chave
Restauração de sistemas de distribuição de energia elétrica, Algoritmo de busca tabu, Representação nó-profundidade, Geração distribuída, Cargas controladas termostaticamente, Controle de carga pelo lado da demanda, Restoration, Real time, Meta-heuritic, Tabu search, Distribution generation, Cold load pick-up, Demand response program